import pandas as pd
import numpy as np

# import data

dt = pd.read_csv("data.csv", sep=";")

dt.head()

we have data redundancies (mois)

dt.shape

(414, 2)

dt.info()

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 414 entries, 0 to 413
Data columns (total 2 columns):
 #   Column             Non-Null Count  Dtype 
---  ------             --------------  ----- 
 0   mois               414 non-null    object
 1   quantites_reelles  414 non-null    int64 
dtypes: int64(1), object(1)
memory usage: 6.6+ KB

dt.mean()

quantites_reelles    2833.379227
dtype: float64

# import covide data

covid19_tn = pd.read_csv("covid19_tn.csv", sep=";")

covid19_tn["date"][1]

'09/03/2020'

we want change the format to yyyy-mm like our data.csv

for i in range(len(covid19_tn)):
    covid19_tn["date"][i]=covid19_tn["date"][i][3:]
    covid19_tn["date"][i]=covid19_tn["date"][i].replace("/",'-')
    covid19_tn["date"][i]=covid19_tn["date"][i][3:]+"-"+covid19_tn["date"][i][:2]

C:\Users\raed\Anaconda3\envs\ML\lib\site-packages\ipykernel_launcher.py:2: SettingWithCopyWarning:


A value is trying to be set on a copy of a slice from a DataFrame

See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy

C:\Users\raed\Anaconda3\envs\ML\lib\site-packages\ipykernel_launcher.py:3: SettingWithCopyWarning:


A value is trying to be set on a copy of a slice from a DataFrame

See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy

C:\Users\raed\Anaconda3\envs\ML\lib\site-packages\ipykernel_launcher.py:4: SettingWithCopyWarning:


A value is trying to be set on a copy of a slice from a DataFrame

See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy

#covid19_tn.to_csv("covid19_tn_For_model.csv",index=False)

covid19_tn.head()

covid19_tn1=covid19_tn.groupby('date').max()
covid19_tn1.reset_index(inplace=True)
covid19_tn1 = covid19_tn1.rename(columns = {'index':'date'})
covid19_tn1=covid19_tn1.rename(columns = {'date': 'mois'}, inplace = False)
covid19_tn1

covid19_tn1.info()

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 9 entries, 0 to 8
Data columns (total 2 columns):
 #   Column  Non-Null Count  Dtype  
---  ------  --------------  -----  
 0   mois    9 non-null      object 
 1   case    9 non-null      float64
dtypes: float64(1), object(1)
memory usage: 272.0+ bytes

covid19_tn1['mois']=covid19_tn1['mois'].apply(str)

import seaborn as sns 
import matplotlib.pyplot as plt 
#g1 = plt.subplot(212)
g1 = sns.pointplot(x="mois", y ="case",data=covid19_tn1)
g1.set_xlabel("mois", fontsize=12)
g1.set_ylabel("case", fontsize=12)
g1.set_title("covide-16 tunisie", fontsize=20)

plt.subplots_adjust(wspace = 0.5, hspace = 0.5,top = 1)

plt.show()

import plotly.graph_objs as go

import plotly.graph_objs as go 
import plotly.tools as tls # It's useful to we get some tools of plotly
#covid19_tn1['mois'] = list(covid19_tn1.index)
fig = go.Figure()
fig.add_trace(go.Scatter(x=covid19_tn1['mois'], y=covid19_tn1['case'],
                    mode='lines+markers',name='Total Cases'))
fig.update_layout(title_text='Confirmed cases in Tunisia each MONTH',plot_bgcolor='rgb(250, 242, 242)')
fig.show()

we have to sum the "quantites_reelles" group by mois

df=dt.groupby('mois').sum()

df.reset_index(inplace=True)
df = df.rename(columns = {'index':'mois'})

df

merging the two data

df_finale=df.merge(covid19_tn1, on = ['mois'],how = 'outer')

df_finale

df_finale.to_csv("df_finale.csv",index=False)

Prediction using Machine Learning Models¶¶

df.merge(covid19_tn1, on = ['mois'])

Using machine learning techniques and with this dataset it's impossible to make a prediction here we can use GANs algorithm to generate more data ,but let's imagine that we have a big dataset and let's make some work

I don't want to add random values in quantity because they will augment the loss then i choose to ignore the column because i don't have A real data

We will use covid19_tn to make a predictions

covid19_tn_model = pd.read_csv("covid19_tn.csv", sep=";")

covid19_tn_model

import plotly.express as px
fig=px.bar(x=covid19_tn_model["date"],y=covid19_tn_model["case"])
fig.update_layout(title="Distribution of Number of  Cases",
                  xaxis_title="Date",yaxis_title="Number of Cases")
fig.show()

fig=go.Figure()
fig.add_trace(go.Scatter(x=covid19_tn_model["date"],y=covid19_tn_model["case"],
                    mode='lines+markers',
                    name='Confirmed Cases'))

fig.update_layout(title="Number cases",
                 xaxis_title="Date",yaxis_title="Number of Cases",legend=dict(x=0,y=1,traceorder="normal"))
fig.show()

Linear Regression Model for Confirm Cases Prediction¶

from sklearn.linear_model import LinearRegression,Ridge,Lasso
from sklearn.metrics import mean_squared_error,r2_score

train_ml=covid19_tn_model.iloc[:int(covid19_tn_model.shape[0]*0.95)]
valid_ml=covid19_tn_model.iloc[int(covid19_tn_model.shape[0]*0.95):]
model_scores=[]

lin_reg=LinearRegression(normalize=True)

lin_reg.fit(np.array(train_ml.index).reshape(-1,1),np.array(train_ml["case"]).reshape(-1,1))

LinearRegression(copy_X=True, fit_intercept=True, n_jobs=None, normalize=True)

prediction_valid_linreg=lin_reg.predict(np.array(valid_ml.index).reshape(-1,1))

model_scores.append(np.sqrt(mean_squared_error(valid_ml["case"],prediction_valid_linreg)))
print("Root Mean Square Error for Linear Regression: ",np.sqrt(mean_squared_error(valid_ml["case"],prediction_valid_linreg)))

Root Mean Square Error for Linear Regression:  51886.249599323215

plt.figure(figsize=(11,6))
prediction_linreg=lin_reg.predict(np.array(covid19_tn_model.index).reshape(-1,1))
linreg_output=[]
for i in range(prediction_linreg.shape[0]):
    linreg_output.append(prediction_linreg[i][0])

fig=go.Figure()
fig.add_trace(go.Scatter(x=covid19_tn_model["date"],y=covid19_tn_model["case"],
                    mode='lines+markers',name="Train Data for Confirmed Cases"))
fig.add_trace(go.Scatter(x=covid19_tn_model["date"], y=linreg_output,
                    mode='lines',name="Linear Regression Best Fit Line",
                    line=dict(color='black', dash='dot')))
fig.update_layout(title="Confirmed Cases Linear Regression Prediction",
                 xaxis_title="Date",yaxis_title="Confirmed Cases",legend=dict(x=0,y=1,traceorder="normal"))
fig.show()

<Figure size 792x432 with 0 Axes>

The Linear Regression Model is absolutely falling aprat. As it is clearly visible that the trend of Confirmed Cases in absolutely not Linear.

Polynomial Regression for Prediction of Confirmed Cases¶

from sklearn.preprocessing import PolynomialFeatures

poly = PolynomialFeatures(degree = 8)

train_ml=covid19_tn_model.iloc[:int(covid19_tn_model.shape[0]*0.95)]
valid_ml=covid19_tn_model.iloc[int(covid19_tn_model.shape[0]*0.95):]
model_scores=[]

train_poly=poly.fit_transform(np.array(train_ml.index).reshape(-1,1))
valid_poly=poly.fit_transform(np.array(valid_ml.index).reshape(-1,1))
y=train_ml["case"]

linreg=LinearRegression(normalize=True)
linreg.fit(train_poly,y)

LinearRegression(copy_X=True, fit_intercept=True, n_jobs=None, normalize=True)

prediction_poly=linreg.predict(valid_poly)
rmse_poly=np.sqrt(mean_squared_error(valid_ml["case"],prediction_poly))
model_scores.append(rmse_poly)
print("Root Mean Squared Error for Polynomial Regression: ",rmse_poly)

Root Mean Squared Error for Polynomial Regression:  11778.321488639067

comp_data=poly.fit_transform(np.array(covid19_tn_model.index).reshape(-1,1))
plt.figure(figsize=(11,6))
predictions_poly=linreg.predict(comp_data)

fig=go.Figure()
fig.add_trace(go.Scatter(x=covid19_tn_model["date"], y=covid19_tn_model["case"],
                    mode='lines+markers',name="Train Data for Confirmed Cases"))
fig.add_trace(go.Scatter(x=covid19_tn_model["date"], y=predictions_poly,
                    mode='lines',name="Polynomial Regression Best Fit",
                    line=dict(color='black', dash='dot')))
fig.update_layout(title="Confirmed Cases Polynomial Regression Prediction",
                 xaxis_title="Date",yaxis_title="Confirmed Cases",
                 legend=dict(x=0,y=1,traceorder="normal"))
fig.show()

<Figure size 792x432 with 0 Axes>

i dont like this model

Support Vector Machine ModelRegressor for Prediction of Confirmed Cases¶¶

from sklearn.svm import SVR

train_ml=covid19_tn_model.iloc[:int(covid19_tn_model.shape[0]*0.95)]
valid_ml=covid19_tn_model.iloc[int(covid19_tn_model.shape[0]*0.95):]
model_scores=[]

#Intializing SVR Model
svm=SVR(C=1,degree=6,kernel='poly',epsilon=0.01)

#Fitting model on the training data
svm.fit(np.array(train_ml.index).reshape(-1,1),np.array(train_ml["case"]).reshape(-1,1))

C:\Users\raed\Anaconda3\envs\ML\lib\site-packages\sklearn\utils\validation.py:760: DataConversionWarning:

A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().

SVR(C=1, cache_size=200, coef0=0.0, degree=6, epsilon=0.01, gamma='scale',
    kernel='poly', max_iter=-1, shrinking=True, tol=0.001, verbose=False)

prediction_valid_svm=svm.predict(np.array(valid_ml.index).reshape(-1,1))

model_scores.append(np.sqrt(mean_squared_error(valid_ml["case"],prediction_valid_svm)))
print("Root Mean Square Error for Support Vectore Machine: ",np.sqrt(mean_squared_error(valid_ml["case"],prediction_valid_svm)))

Root Mean Square Error for Support Vectore Machine:  7251.9351697814445

plt.figure(figsize=(11,6))
prediction_svm=svm.predict(np.array(covid19_tn_model.index).reshape(-1,1))
fig=go.Figure()
fig.add_trace(go.Scatter(x=covid19_tn_model["date"], y=covid19_tn_model["case"],
                    mode='lines+markers',name="Train Data for Confirmed Cases"))
fig.add_trace(go.Scatter(x=covid19_tn_model["date"], y=prediction_svm,
                    mode='lines',name="Support Vector Machine Best fit Kernel",
                    line=dict(color='black', dash='dot')))
fig.update_layout(title="Confirmed Cases Support Vectore Machine Regressor Prediction",
                 xaxis_title="Date",yaxis_title="Confirmed Cases",legend=dict(x=0,y=1,traceorder="normal"))
fig.show()

<Figure size 792x432 with 0 Axes>

as you can see SVM is the best model in our prediction¶

Time Series Forecasting¶

let's begin with Holt's Linear Model¶

from statsmodels.tsa.api import Holt,SimpleExpSmoothing,ExponentialSmoothing

train_ml=covid19_tn_model.iloc[:int(covid19_tn_model.shape[0]*0.95)]
valid=covid19_tn_model.iloc[int(covid19_tn_model.shape[0]*0.95):]
y_pred=valid.copy()

holt=Holt(np.asarray(train_ml["case"])).fit(smoothing_level=0.4, smoothing_slope=0.4,optimized=False)

y_pred["Holt"]=holt.forecast(len(valid))
model_scores.append(np.sqrt(mean_squared_error(y_pred["case"],y_pred["Holt"])))
print("Root Mean Square Error Holt's Linear Model: ",np.sqrt(mean_squared_error(y_pred["case"],y_pred["Holt"])))

Root Mean Square Error Holt's Linear Model:  2014.1872133271897

fig=go.Figure()
fig.add_trace(go.Scatter(x=covid19_tn_model.index, y=covid19_tn_model["case"],
                    mode='lines+markers',name="Train Data for Confirmed Cases"))
fig.add_trace(go.Scatter(x=valid.index, y=valid["case"],
                    mode='lines+markers',name="Validation Data for Confirmed Cases",))
fig.add_trace(go.Scatter(x=valid.index, y=y_pred["Holt"],
                    mode='lines+markers',name="Prediction of Confirmed Cases",))
fig.update_layout(title="Confirmed Cases Holt's Linear Model Prediction",
                 xaxis_title="Date",yaxis_title="Confirmed Cases",legend=dict(x=0,y=1,traceorder="normal"))
fig.show()

AR Model (using AUTO ARIMA)¶

from pmdarima.arima import auto_arima
import pmdarima as pm

#pip install pmdarima   ----- try this

model_train=covid19_tn_model.iloc[:int(covid19_tn_model.shape[0]*0.95)]
valid=covid19_tn_model.iloc[int(covid19_tn_model.shape[0]*0.95):]
y_pred=valid.copy()

model_ar= auto_arima(covid19_tn_model["case"],trace=True, error_action='ignore', start_p=0,start_q=0,max_p=4,max_q=0,
                   suppress_warnings=True,stepwise=False,seasonal=False)
model_ar.fit(covid19_tn_model["case"])

 ARIMA(0,2,0)(0,0,0)[0] intercept   : AIC=4311.409, Time=1.08 sec
 ARIMA(1,2,0)(0,0,0)[0] intercept   : AIC=4194.484, Time=0.06 sec
 ARIMA(2,2,0)(0,0,0)[0] intercept   : AIC=4109.242, Time=0.21 sec
 ARIMA(3,2,0)(0,0,0)[0] intercept   : AIC=4103.417, Time=0.10 sec
 ARIMA(4,2,0)(0,0,0)[0] intercept   : AIC=4095.277, Time=0.10 sec
Total fit time: 1.591 seconds

ARIMA(maxiter=50, method='lbfgs', order=(4, 2, 0), out_of_sample_size=0,
      scoring='mse', scoring_args={}, seasonal_order=(0, 0, 0, 0),
      start_params=None, suppress_warnings=True, trend=None,
      with_intercept=True)

prediction_ar=model_ar.predict(len(valid))
y_pred["AR Model Prediction"]=prediction_ar

model_scores.append(np.sqrt(mean_squared_error(y_pred["case"],y_pred["AR Model Prediction"])))
print("Root Mean Square Error for AR Model: ",np.sqrt(mean_squared_error(y_pred["case"],y_pred["AR Model Prediction"])))

Root Mean Square Error for AR Model:  15084.032144596507

fig=go.Figure()
fig.add_trace(go.Scatter(x=model_train.index, y=model_train["case"],
                    mode='lines+markers',name="Train Data for Confirmed Cases"))
fig.add_trace(go.Scatter(x=valid.index, y=valid["case"],
                    mode='lines+markers',name="Validation Data for Confirmed Cases",))
fig.add_trace(go.Scatter(x=valid.index, y=y_pred["AR Model Prediction"],
                    mode='lines+markers',name="Prediction of Confirmed Cases",))
fig.update_layout(title="Confirmed Cases AR Model Prediction",
                 xaxis_title="Date",yaxis_title="Confirmed Cases",legend=dict(x=0,y=1,traceorder="normal"))
fig.show()

not good¶

MA Model (using AUTO ARIMA)¶

model_train=covid19_tn_model.iloc[:int(covid19_tn_model.shape[0]*0.95)]
valid=covid19_tn_model.iloc[int(covid19_tn_model.shape[0]*0.95):]
y_pred=valid.copy()

model_ma= auto_arima(model_train["case"],trace=True, error_action='ignore', start_p=0,start_q=0,max_p=0,max_q=2,
                   suppress_warnings=True,stepwise=False,seasonal=False)
model_ma.fit(model_train["case"])

 ARIMA(0,2,0)(0,0,0)[0] intercept   : AIC=4090.775, Time=0.02 sec
 ARIMA(0,2,1)(0,0,0)[0] intercept   : AIC=3874.683, Time=0.26 sec
 ARIMA(0,2,2)(0,0,0)[0] intercept   : AIC=3842.893, Time=0.31 sec
Total fit time: 0.604 seconds

ARIMA(maxiter=50, method='lbfgs', order=(0, 2, 2), out_of_sample_size=0,
      scoring='mse', scoring_args={}, seasonal_order=(0, 0, 0, 0),
      start_params=None, suppress_warnings=True, trend=None,
      with_intercept=True)

prediction_ma=model_ma.predict(len(valid))
y_pred["MA Model Prediction"]=prediction_ma

model_scores.append(np.sqrt(mean_squared_error(valid["case"],prediction_ma)))
print("Root Mean Square Error for MA Model: ",np.sqrt(mean_squared_error(valid["case"],prediction_ma)))

Root Mean Square Error for MA Model:  2217.486447676189

fig=go.Figure()
fig.add_trace(go.Scatter(x=model_train.index, y=model_train["case"],
                    mode='lines+markers',name="Train Data for Confirmed Cases"))
fig.add_trace(go.Scatter(x=valid.index, y=valid["case"],
                    mode='lines+markers',name="Validation Data for Confirmed Cases",))
fig.add_trace(go.Scatter(x=valid.index, y=y_pred["MA Model Prediction"],
                    mode='lines+markers',name="Prediction for Confirmed Cases",))
fig.update_layout(title="Confirmed Cases MA Model Prediction",
                 xaxis_title="Date",yaxis_title="Confirmed Cases",legend=dict(x=0,y=1,traceorder="normal"))
fig.show()

goood model¶

ARIMA Model (using AUTOARIMA)¶

model_train=covid19_tn_model.iloc[:int(covid19_tn_model.shape[0]*0.95)]
valid=covid19_tn_model.iloc[int(covid19_tn_model.shape[0]*0.95):]
y_pred=valid.copy()

model_arima= auto_arima(model_train["case"],trace=True, error_action='ignore', start_p=1,start_q=1,max_p=3,max_q=3,
                   suppress_warnings=True,stepwise=False,seasonal=False)
model_arima.fit(model_train["case"])

 ARIMA(0,2,0)(0,0,0)[0] intercept   : AIC=4090.775, Time=0.03 sec
 ARIMA(0,2,1)(0,0,0)[0] intercept   : AIC=3874.683, Time=0.21 sec
 ARIMA(0,2,2)(0,0,0)[0] intercept   : AIC=3842.893, Time=0.26 sec
 ARIMA(0,2,3)(0,0,0)[0] intercept   : AIC=3822.394, Time=0.52 sec
 ARIMA(1,2,0)(0,0,0)[0] intercept   : AIC=3977.330, Time=0.04 sec
 ARIMA(1,2,1)(0,0,0)[0] intercept   : AIC=3849.577, Time=0.41 sec
 ARIMA(1,2,2)(0,0,0)[0] intercept   : AIC=3810.224, Time=0.53 sec
 ARIMA(1,2,3)(0,0,0)[0] intercept   : AIC=3821.494, Time=0.41 sec
 ARIMA(2,2,0)(0,0,0)[0] intercept   : AIC=3892.401, Time=0.06 sec
 ARIMA(2,2,1)(0,0,0)[0] intercept   : AIC=3843.314, Time=0.37 sec
 ARIMA(2,2,2)(0,0,0)[0] intercept   : AIC=3819.415, Time=0.40 sec
 ARIMA(2,2,3)(0,0,0)[0] intercept   : AIC=3820.987, Time=0.53 sec
 ARIMA(3,2,0)(0,0,0)[0] intercept   : AIC=3886.762, Time=0.07 sec
 ARIMA(3,2,1)(0,0,0)[0] intercept   : AIC=3842.477, Time=0.52 sec
 ARIMA(3,2,2)(0,0,0)[0] intercept   : AIC=3821.019, Time=0.88 sec
Total fit time: 5.263 seconds

ARIMA(maxiter=50, method='lbfgs', order=(1, 2, 2), out_of_sample_size=0,
      scoring='mse', scoring_args={}, seasonal_order=(0, 0, 0, 0),
      start_params=None, suppress_warnings=True, trend=None,
      with_intercept=True)

prediction_arima=model_arima.predict(len(valid))
y_pred["ARIMA Model Prediction"]=prediction_arima

model_scores.append(np.sqrt(mean_squared_error(valid["case"],prediction_arima)))
print("Root Mean Square Error for ARIMA Model: ",np.sqrt(mean_squared_error(valid["case"],prediction_arima)))

Root Mean Square Error for ARIMA Model:  1133.617304492231

fig=go.Figure()
fig.add_trace(go.Scatter(x=model_train.index, y=model_train["case"],
                    mode='lines+markers',name="Train Data for Confirmed Cases"))
fig.add_trace(go.Scatter(x=valid.index, y=valid["case"],
                    mode='lines+markers',name="Validation Data for Confirmed Cases",))
fig.add_trace(go.Scatter(x=valid.index, y=y_pred["ARIMA Model Prediction"],
                    mode='lines+markers',name="Prediction for Confirmed Cases",))
fig.update_layout(title="Confirmed Cases ARIMA Model Prediction",
                 xaxis_title="Date",yaxis_title="Confirmed Cases",legend=dict(x=0,y=1,traceorder="normal"))
fig.show()

SARIMA Model (using AUTO ARIMA)¶

model_sarima= auto_arima(model_train["case"],trace=True, error_action='ignore', 
                         start_p=0,start_q=0,max_p=2,max_q=2,m=7,
                   suppress_warnings=True,stepwise=True,seasonal=True)
model_sarima.fit(model_train["case"])

Performing stepwise search to minimize aic
 ARIMA(0,2,0)(1,0,1)[7]             : AIC=4070.373, Time=0.20 sec
 ARIMA(0,2,0)(0,0,0)[7]             : AIC=4088.796, Time=0.04 sec
 ARIMA(1,2,0)(1,0,0)[7]             : AIC=3968.481, Time=0.13 sec
 ARIMA(0,2,1)(0,0,1)[7]             : AIC=3859.356, Time=0.34 sec
 ARIMA(0,2,1)(0,0,0)[7]             : AIC=3876.171, Time=0.09 sec
 ARIMA(0,2,1)(1,0,1)[7]             : AIC=3861.200, Time=0.46 sec
 ARIMA(0,2,1)(0,0,2)[7]             : AIC=3861.317, Time=0.87 sec
 ARIMA(0,2,1)(1,0,0)[7]             : AIC=3860.242, Time=0.29 sec
 ARIMA(0,2,1)(1,0,2)[7]             : AIC=inf, Time=1.13 sec
 ARIMA(0,2,0)(0,0,1)[7]             : AIC=4067.259, Time=0.13 sec
 ARIMA(1,2,1)(0,0,1)[7]             : AIC=3841.472, Time=0.24 sec
 ARIMA(1,2,1)(0,0,0)[7]             : AIC=3851.237, Time=0.12 sec
 ARIMA(1,2,1)(1,0,1)[7]             : AIC=3840.714, Time=0.90 sec
 ARIMA(1,2,1)(1,0,0)[7]             : AIC=3841.931, Time=0.19 sec
 ARIMA(1,2,1)(2,0,1)[7]             : AIC=inf, Time=1.46 sec
 ARIMA(1,2,1)(1,0,2)[7]             : AIC=inf, Time=1.46 sec
 ARIMA(1,2,1)(0,0,2)[7]             : AIC=3843.307, Time=0.61 sec
 ARIMA(1,2,1)(2,0,0)[7]             : AIC=3843.904, Time=0.40 sec
 ARIMA(1,2,1)(2,0,2)[7]             : AIC=3838.119, Time=1.23 sec
 ARIMA(0,2,1)(2,0,2)[7]             : AIC=3860.688, Time=1.40 sec
 ARIMA(1,2,0)(2,0,2)[7]             : AIC=3968.780, Time=0.84 sec
 ARIMA(2,2,1)(2,0,2)[7]             : AIC=3830.364, Time=1.58 sec
 ARIMA(2,2,1)(1,0,2)[7]             : AIC=inf, Time=1.46 sec
 ARIMA(2,2,1)(2,0,1)[7]             : AIC=inf, Time=1.22 sec
 ARIMA(2,2,1)(1,0,1)[7]             : AIC=3832.477, Time=0.81 sec
 ARIMA(2,2,0)(2,0,2)[7]             : AIC=inf, Time=1.08 sec
 ARIMA(2,2,2)(2,0,2)[7]             : AIC=inf, Time=1.88 sec
 ARIMA(1,2,2)(2,0,2)[7]             : AIC=inf, Time=1.66 sec
 ARIMA(2,2,1)(2,0,2)[7] intercept   : AIC=3835.359, Time=1.53 sec

Best model:  ARIMA(2,2,1)(2,0,2)[7]          
Total fit time: 24.021 seconds

ARIMA(maxiter=50, method='lbfgs', order=(2, 2, 1), out_of_sample_size=0,
      scoring='mse', scoring_args={}, seasonal_order=(2, 0, 2, 7),
      start_params=None, suppress_warnings=True, trend=None,
      with_intercept=False)

prediction_sarima=model_sarima.predict(len(valid))
y_pred["SARIMA Model Prediction"]=prediction_sarima

model_scores.append(np.sqrt(mean_squared_error(y_pred["case"],y_pred["SARIMA Model Prediction"])))
print("Root Mean Square Error for SARIMA Model: ",np.sqrt(mean_squared_error(y_pred["case"],y_pred["SARIMA Model Prediction"])))

Root Mean Square Error for SARIMA Model:  2457.0755041337648

fig=go.Figure()
fig.add_trace(go.Scatter(x=model_train.index, y=model_train["case"],
                    mode='lines+markers',name="Train Data for Confirmed Cases"))
fig.add_trace(go.Scatter(x=valid.index, y=valid["case"],
                    mode='lines+markers',name="Validation Data for Confirmed Cases",))
fig.add_trace(go.Scatter(x=valid.index, y=y_pred["SARIMA Model Prediction"],
                    mode='lines+markers',name="Prediction for Confirmed Cases",))
fig.update_layout(title="Confirmed Cases SARIMA Model Prediction",
                 xaxis_title="Date",yaxis_title="Confirmed Cases",legend=dict(x=0,y=1,traceorder="normal"))
fig.show()

feedback information¶

SVM AND ARIMA ARE THE BEST MODELS

	mois	quantites_reelles
0	2020-02	1352
1	2020-03	2016
2	2020-04	1179
3	2020-05	601
4	2018-01	1087

	mois	case
0	2020-03	312.0
1	2020-04	980.0
2	2020-05	1076.0
3	2020-06	1172.0
4	2020-07	1514.0
5	2020-08	3685.0
6	2020-09	17405.0
7	2020-10	59813.0
8	2020-11	92475.0

	mois	quantites_reelles
0	2018-01	38731
1	2018-02	35837
2	2018-03	41470
3	2018-04	42260
4	2018-05	44181
5	2018-06	37538
6	2018-07	35772
7	2018-08	39691
8	2018-09	64329
9	2018-10	56186
10	2018-11	36473
11	2018-12	28906
12	2019-01	62317
13	2019-02	40296
14	2019-03	44123
15	2019-04	40335
16	2019-05	42041
17	2019-06	27989
18	2019-07	26646
19	2019-08	33304
20	2019-09	52272
21	2019-10	35119
22	2019-11	21722
23	2019-12	23927
24	2020-01	48463
25	2020-02	60080
26	2020-03	79955
27	2020-04	31613
28	2020-05	1443

	mois	quantites_reelles	case
0	2018-01	38731.0	NaN
1	2018-02	35837.0	NaN
2	2018-03	41470.0	NaN
3	2018-04	42260.0	NaN
4	2018-05	44181.0	NaN
5	2018-06	37538.0	NaN
6	2018-07	35772.0	NaN
7	2018-08	39691.0	NaN
8	2018-09	64329.0	NaN
9	2018-10	56186.0	NaN
10	2018-11	36473.0	NaN
11	2018-12	28906.0	NaN
12	2019-01	62317.0	NaN
13	2019-02	40296.0	NaN
14	2019-03	44123.0	NaN
15	2019-04	40335.0	NaN
16	2019-05	42041.0	NaN
17	2019-06	27989.0	NaN
18	2019-07	26646.0	NaN
19	2019-08	33304.0	NaN
20	2019-09	52272.0	NaN
21	2019-10	35119.0	NaN
22	2019-11	21722.0	NaN
23	2019-12	23927.0	NaN
24	2020-01	48463.0	NaN
25	2020-02	60080.0	NaN
26	2020-03	79955.0	312.0
27	2020-04	31613.0	980.0
28	2020-05	1443.0	1076.0
29	2020-06	NaN	1172.0
30	2020-07	NaN	1514.0
31	2020-08	NaN	3685.0
32	2020-09	NaN	17405.0
33	2020-10	NaN	59813.0
34	2020-11	NaN	92475.0

	mois	quantites_reelles	case
0	2020-03	79955	312.0
1	2020-04	31613	980.0
2	2020-05	1443	1076.0

	date	case
0	03/03/2020	1.0
1	09/03/2020	1.0
2	10/03/2020	2.0
3	11/03/2020	5.0
4	12/03/2020	7.0
...	...	...
260	23/11/2020	88711.0
261	24/11/2020	89195.0
262	25/11/2020	90213.0
263	26/11/2020	91307.0
264	27/11/2020	92475.0